[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: color temperature puzzle

At 10:16 AM 9/1/98, Tarcis Verfaillie wrote:
>Dear all,
>I have a question that is keeping me puzzled.
>Why is there a difference between the colortemperature of a monitor (measured
>correctly with a dedicated color temperature meter/probe) and for example light
>on the set measured with a light colortemperature???
>In other words:  I measure the colortemperature with the CE 100 from Minolta,
>which is the dedicated for monitor calibration, it reads on my monitor 5600K
>(example).  When I then dubble check with a Minolta Colormeter II or III it
>reads about 1000K less (4600K)!!!  So we are fooled here!!!  I called Minolta
>and they were telling me that the standard for both meters have been set by the
>CIE in France.  Looking into that I learned that these standards have been set
>even before colorTV was worhtwhile looking at....
>Is there something I do not understand or do we as colorists have to write to
>the CIE look into the standards???????  Can anybody help me on this topic.

Because of the nature of the phosphors which emit light in a monitor,
setting monitor color is both difficult and very misleading to a colorist
when set incorrectly.

There are two different but related problems with your attempt to measure
color temperature of a monitor and then compare this with the color
temperature of the illumination on a set.

The first is that, strictly speaking, the monitor does not have a color
temperature.  Color temperature refers to the radiation emitted by a
theoretical black body radiator.  Incandescent lamps with tungsten
filaments closely approximate a black body radiator, with light being
predictably emitted throughout the visible part of the electromagnetic
spectrum and therefore,  can  be measured with a simple color temperature
meter which compares the amount of "red" and the amount of "blue" light.

A monitor contains three phosphors which emit "Blue", "green", and "red"
wavelengths discontinuously, i.e.,  there are gaps in the spectrum of light
emitted which confuse a simple color temperature meter( and sometimes
complex and costly meters!).  When referring to the color temperature of
monitors, HMI lights, fluorescent lamps and other discontinuous light
sources the correct term is "correlated color temperature".  That is, the
color temperature of a black body radiator which would most closely match
the visual appearance of the discontinuous spectrum source.

Which brings us to the second problem.  Color is not a physical property of
materials, but rather the human visual systems response to electromagnetic
radiation of wavelengths from approximately 400 to 700nm.  Therefore any
device which would measure color would need to mimic the human eye response
to light.  This has proven to be surprisingly difficult to do. The only
reliable method I know of, is to use a spectrometer with at least 2nm
resolution and software which will calculate the resulting correlated color
temperature.  Most, so called, color temperature meters use 2, 3 and
sometimes 4 sensors to measure color(my spectrometer measures 2048
wavelengths from 380 to 1100nm).

I have compared two devices for measuring monitor color temperature which
give results which compare with my spectrometer, one is made by Philips
Electronics and the other by Grasby Electronics.  I am sorry but I do not
have model numbers or contacts at this time.

John Galt,
Vice President High Definition Technology Development
Sony Pictures High Definition Center
10202 West Washington Blvd.
Culver City
CA 90232-3195

VOICE: 310-244-4387     FAX: 310-244-3014
E-MAIL: johngalt at earthlink.net

Thanks to Richard Cassel for support in 1998.
No product marketing allowed on the main TIG.  Contact rob at alegria.com
1012 subscribers in 39 countries on Fri Sep  4 21:40:46 CDT 1998 
subscribe/unsubscribe with that Subject: to telecine-request at alegria.com
complete information on the TIG website http://www.alegria.com/tig3/